Re-scans your project’s tech stack and runs the AI enhancer to update the codebase context in .ralph/.context.json. Use this after significant changes to your project — new dependencies, restructured directories, or updated conventions.
What it does
- Scans — Re-runs the full tech stack detection (same detectors as
wiggum init)
- Analyzes — Runs the AI enhancer to update entry points, key directories, commands, naming conventions, and implementation guidelines
- Persists — Saves the updated context to
.ralph/.context.json
The updated context feeds into all subsequent spec generation (wiggum new) and loop execution (wiggum run).
When to use
- After adding major new dependencies or frameworks
- After restructuring your project directory layout
- After changing build/test/lint commands
- When specs or loops seem to produce code that doesn’t match your current patterns
- Periodically, to keep context fresh
wiggum sync does not re-run the full init flow — it skips provider selection, API key setup, and file generation. It only updates the codebase analysis.
Examples
